Zoomcamp MLOps Course

More details

1. Introduction

• 1.1 What is MLOps
• 1.2 MLOps maturity model
• 1.3 Running example: NY Taxi trips dataset
• 1.4 Why do we need MLOps
• 1.5 Environment preparation

2. Experiment tracking and model management

• 2.1 Experiment tracking intro
• 2.2 Getting started with MLflow
• 2.3 Experiment tracking with MLflow
• 2.4 Saving and loading models with MLflow
• 2.5 Model registry
• 2.6 MLflow in practice

3. Orchestration and ML Pipelines

• 3.1 Workflow orchestration
• 3.2 Mage

4. Model Deployment

• 4.1 Three ways of model deployment: Online (web and streaming) and offline (batch)
• 4.2 Web service: model deployment with Flask
• 4.3 Streaming: consuming events with AWS Kinesis and Lambda
• 4.4 Batch: scoring data offline

5. Model Monitoring

• 5.1 Monitoring ML-based services
• 5.2 Monitoring web services with Prometheus, Evidently, and Grafana
• 5.3 Monitoring batch jobs with Prefect, MongoDB, and Evidently

6. Best Practices

• 6.1 Testing: unit, integration
• 6.2 Python: linting and formatting
• 6.3 Pre-commit hooks and makefiles
• 6.4 CI/CD (GitHub Actions)
• 6.5 Infrastructure as code (Terraform)

7. Project

• 7.1 End-to-end project with all the things above